Background

Three-dimensional (3D) cell culture methods are widely accepted as being more physiologically relevant than conventional two-dimensional (2D)

methods. Cellular functions and responses that exist in tissues are often lost in 2D cell cultures, limiting their predictive capability for drug efficacy screening [Pampaloni et al., 2007; Windus et al., 2012]. Microtissues, or spheroids, are one of the most well characterized models for 3D culture and cell-based drug screening, due to their reproducibility and similarity to tissues in an organism. Microtissues are self-assembled spherical clusters of cells cultured in environments where cell-cell interactions dominate over cell-substrate interactions. Microtissues consisting of cancer cells naturally mimic avascular tumors, with inherent metabolic (oxygen) and proliferative (nutrient) gradients [Friedrich et al., 2009], and therefore serve as excellent physiological tumor models.

Cellular Imaging and Analysis

A P P L I C A T I O N N O T E

Key Features

• High quality three-dimensional imaging of microtissues

• Significantly increase detection depth into the core of microtissues

Imaging Microtissue Cores using the Opera High Content Screening System

Tissue Core Signal Intensity

Optical Clearing

2

Since cancer microtissues develop different cell layers, typically a proliferative rim, a quiescent region and a hypoxic or necrotic core, it is of great interest to image all regions into the center of the tissues. However, visualization of 3D microtissues using conventional confocal microscopy is challenging, because tissue opacity, light absorption and scattering prevent imaging at higher planes deep into the core of microtissues. The introduction of novel methods for optical clearing of tissues represents a major advance in the field [Ertürk et al., 2012; Hama et al., 2011]. By combining the high quality optics of the Opera® High Content Screening System (equipped with a water immersion lens) with an optical clearance solution, we are pushing the boundaries of microtissue imaging. Pre-treatment of microtissues with the aqueous reagent Scale renders tissues optically more transparent, while fully preserving fluorescent signals in the clarified structures. This doubles the detection depth, thus allowing more detailed analysis of all cell layers within cancer microtissues.

Application

HCT116 microtissues (human colon carcinoma cell line stably expressing cytosolic GFP) with a cell seeding density of 750 cells per microtissue, and NIH3T3 microtissues (mouse embryonic fibroblast cell line stably expressing cytosolic RFP) with a seeding density of 3000 cells per microtissue, were produced using the GravityPLUS™ system from InSphero AG.

Figure 1: Image acquisition and analysis of untreated and Scale-treated microtissues. (A) Stacks of 100 images of untreated and Scale-treated tissues, with a 2 µm plane distance and a total z-height of 200 µm, were acquired in the Hoechst, GFP and RFP channels using the Opera system equipped with a 20x W objective. (B) Using the Acapella software, all planes of the Hoechst channel (upper panels) were analyzed using the Find Nuclei building block (mid panels). Texture properties (SER Edge, 6 px) of the microtissue core (lower panels, white circle) were also calculated for each plane and channel using the Calculate Texture Properties building block. Only a selected plane at z=100 µm is shown in all panels.

This system allows for automation of the hanging drop method and provides an optimal tool for making microtissues amenable to cell-based drug screening [Drewitz et al., 2011].

Microtissues were fixed and stained for 1 h with 16 µM Hoechst. The microtissues were then transferred to ViewPlate™-96 glass bottom plate (PerkinElmer #6005430) and either left untreated in PBS or subjected to treatment with Scale reagent composed of 4 M urea, 10% (wt/vol) glycerol and 0.1% (wt/vol) Triton X-100 for 1 day [Hama et al., 2011]. After 24 h incubation at room temperature, tissues were imaged in the presence of the Scale reagent with the Opera system, using a 20x water objective. A 200 µm z-stack with a plane distance of 2 µm was acquired (Fig 1A).

At 100 µm imaging depth, the opacity of light in untreated tissues is already clearly visible. Due to the spherical shape of the tissue, light absorption and scattering by cells from lower planes causes a distinct signal intensity decrease in the center of this optical section. For this reason, untreated tissues appear to be hollow inside (Fig 1B). On the contrary, Scale treatment renders the tissues transparent so as the nuclei in the center of this optical section, at 100 µm, are still clearly visible. To obtain a quantitative measurement of how far into the microtissue core structures were visible, nuclear segmentation was performed and SER Edge texture properties were calculated for each plane using the Acapella® High Content Imaging and Analysis Software (Fig 1B).

200 µm

Untreated Scale-Treated

Input Image

Find Nuclei

Calculate Texture Properties of Microtissue Core

A Bz=100 µm z=100 µmHoechst

100 µm

3

Figure 2: Comparison of texture properties in untreated and Scale-treated HCT116 microtissues. The images show selected planes of the Hoechst channel (A) and GFP channel (B). The graphs show the calculated texture (SER Edge, 6 px) values of the microtissue core for each z plane in the Hoechst channel (C) and the GFP channel (D). Using Acapella software, the texture (SER Edge, 6 px) of the microtissue center (white circle) was analyzed for each z plane. At high z planes (z=100 µm ± 50 µm) , the calculated SER Edge values in the Scale-treated microtissue is significantly higher than in the untreated microtissue. Treatment with Scale therefore increases the detection depth.

To compare texture properties in the different conditions and channels, the SER Edge values were plotted against the z-height (Fig 2). As expected, the calculated texture values decrease with increasing z-height, but at high z planes (z=100 µm ± 50 µm) Scale-treated microtissues show much greater values compared to the untreated controls. Therefore, microtissue imaging benefits from pre-treatment with Scale reagent as more structure becomes visible deeper into the tissue core.

Ultimately, an absolute detection depth was determined for each condition and channel for at least three HCT116 and three NIH3T3 microtissues (Fig 3). The detection depth in the Hoechst channel is derived from the highest image plane suitable for nuclei segmentation in the microtissue core. This Hoechst image plane showed a SER Edge texture value of 0.006 in the microtissue core.

This value was subsequently used as a threshold to determine the detection depth in the GFP channel (HTC116 microtissues) and the RFP channel (NIH3T3 microtissues). Figure 3 illustrates two important effects. On the one hand, Scale treatment doubles the detection depth in all channels and is therefore a very useful tool for microtissue analysis. On the other hand, it is clear that microtissue imaging benefits from dyes with longer excitation and emission wavelengths. When comparing the detection depths in the Hoechst and GFP or Hoechst and RFP channels, GFP and RFP show higher values, as longer wavelength light more readily passes through tissues.

Although the initial seeding density of the NIH3T3 cells was 4 times higher, the microtissues only reached an overall size of 280 µm in diameter, compared to 320 µm for the HCT116 tissues. This is because fibroblasts are contact inhibited and, when cultured as 3D microtissues, divide to a much lesser extent than HCT116 cancer cells. Therefore, the NIH3T3 tissues are more compact and the detection depth is about 20 % lower than in the HCT116 tissues.

200 µm

A

C

B

D

z = 50 µmz = 50 µm

Scale Treatment

Untreated

z = 100 µmz = 100 µm

Hoechst GFP

100 µm100 µm

4

Figure 3: Detection depth in HCT116 (left) and NIH3T3 (right) microtissues (MT’s). The detection depth is determined using SER Edge texture values as a measure for structures in the tissue core region. An appropriate texture threshold to determine the detection depth was derived from the z-height [µm] of the highest image plane suitable for nuclei segmentation in the Hoechst channel. Using the Opera system with the 20x W objective, untreated HCT116 microtissues could be imaged up to 53 µm in the Hoechst channel and 70 µm in the GFP channel. The Scale treatment doubled the detection depth in HCT116 microtissues to 110 µm for Hoechst and 136 µm for GFP. In NIH3T3 microtissues with a more compact structure, Scale treatment increased the detection depth from 42 µm to 74 µm (Hoechst) and from 61 µm to 117 µm (RFP). n = 3 microtissues.

Conclusions

3D microtissues more accurately reflect in vivo cellular properties and function and are therefore one of the most important cell-based model systems for drug discovery. Although the number of reports describing High Content Screening (HCS) methods for 3D spheroids is increasing, it still remains a challenge to adapt spheroid-based assays to currently available HCS platforms [Tung et al., 2011; Vinci et al., 2012]. Besides 3D quantitative image analysis and data handling, one of the major challenges is 3D imaging itself.

To thoroughly monitor the effects of drugs on the diverse layers of a cancer spheroid, imaging up to the spheroid core is required. However, this is strongly affected by tissue opacity, light absorption and scattering. The technique of optical clearance using the Scale reagent doubles the detection depth in microtissues. By combining Scale tissue clearance with the excellent optical properties of the Opera system equipped with water immersion objectives, imaging of the lower half of the spheroid, and thus all spheroid layers, becomes possible. Optical clearance solutions such as the Scale reagent, therefore, provide a simple and effective method for adapting spheroid-based assays to HCS systems, e.g. the Opera system. This will ultimately allow more information to be obtained from deeper inside physiological relevant 3D cell models.

References

Pampaloni, F., Reynaud, E.G., and Stelzer, E.H.K. (2007): The third dimension bridges the gap between cell culture and live tissue. Nature reviews. Molecular cell biology, 8 (10), 839–45.

Windus, L.C.E., Kiss, D.L., Glover, T., and Avery, V.M. (2012): In vivo biomarker expression patterns are preserved in 3D cultures of Prostate Cancer. Experimental cell research, 318 (19), 2507–19.

Friedrich, J., Seidel, C., Ebner, R., and Kunz-Schughart, L.A. (2009): Spheroid-based drug screen: considerations and practical approach. Nature protocols, 4 (3), 309–24.

Ertürk, A., Becker, K., Jährling, N., Mauch, C.P., Hojer, C.D., Egen, J.G., Hellal, F., Bradke, F., Sheng, M., and Dodt, H.U. (2012): Three-dimensional imaging of solvent-cleared organs using 3DISCO. Nature protocols, 7 (11), 1983–95.

Hama, H., Kurokawa, H., Kawano, H., Ando, R., Shimogori, T., Noda, H., Fukami, K., Sakaue-Sawano, A., and Miyawaki, A. (2011): Scale: a chemical approach for fluorescence imaging and reconstruction of transparent mouse brain. Nature Neuroscience, 14 (11), 1481–8.

Drewitz, M., Helbling, M., Fried, N., Bieri, M., Moritz, W., Lichtenberg, J., and Kelm, J.M. (2011): Towards automated production and drug sensitivity testing using scaffold-free spherical tumor microtissues. Biotechnology Journal, 6 (12), 1488–96.

Vinci, M., Gowan, S., Boxall, F., Patterson, L., Zimmermann, M., Court, W., Lomas, C., Mendiola, M., Hardisson, D., and Eccles, S.A. (2012): Advances in establishment and analysis of three-dimensional tumor spheroid-based functional assays for target validation and drug evaluation. BMC Biology, 10 29.

Tung, Y.-C., Hsiao, A.Y., Allen, S.G., Torisawa, Y., Ho, M., and Takayama, S. (2011): High-throughput 3D spheroid culture and drug testing using a 384 hanging drop array. The Analyst, 136 (3), 473–8.

Authors

Stefan Letzsch

Karin Böttcher

PerkinElmer

Cellular Technologies Germany GmbH

Hamburg, DE

Jens Kelm

InSphero

Zurich, CH

320 µm280 µm

Unt

reat

ed

Unt

reat

ed

Scal

e

Scal

e

Detection depth in HCT116 MT‘s Detection depth in NIH3T3 MT‘s

For a complete listing of our global offices, visit www.perkinelmer.com/ContactUs

Copyright ©2013-2014, PerkinElmer, Inc. All rights reserved. PerkinElmer® is a registered trademark of PerkinElmer, Inc. All other trademarks are the property of their respective owners. 011190A_01

PerkinElmer, Inc. 940 Winter Street Waltham, MA 02451 USA P: (800) 762-4000 or (+1) 203-925-4602www.perkinelmer.com